imbalance in output tube current - fender deluxe reverb

[amplifryer]

Hello, First post from a noob here. Please be gentle. I am getting back into working on tube amps after several years of not. Amplifier in question is my '65 Fender Deluxe Reverb AB763. I noticed that at least one of the power tubes was red-plating. Current assumption is it was biased too hot, I cut it back to -50V and began investigating on the bench. Currently it's biased about -40V putting tube cathode currents at ~30mA.

My concern is the imbalance in the output section. One of the output tubes is doing all the work. Its current increases proportionally with increases in amp signal going up 20-30mA while the other 6V6 increases only a few mA beyond bias current. I understand that some imbalance here is fairly common but this seems excessive to me.

All of the components in the output section are within spec., however the quiescent plate voltages on the PI vary significantly depending on which tube is installed. With a sine wave input, the V6B signal at the plate is over twice the amplitude of V6A. For the values shown on the diagram, I used a JJ ECC83S in the PI position as it had the best plate voltage balance. I'll replace it with a new 12AT7, but the ones I have are pretty lopsided.

When driven hard, the output waveform clips on one polarity well before the opposite. Does this seem normal? I don't have another amp to compare. I know I can balance the signals by experimenting with LTPI resistor values, but I don't want to modify something that needs to be fixed first.

[wpaulvogel]

What happens if you swap the 6V6 tubes between the sockets. Watch and record the current again.

[xtian]

Is this a vintage 1965 amp or a reissue? If it's really 1965, I'd be concerned the 0.1u coupling caps are leaky.

[amplifryer]

Thanks for the responses guys. This is a vintage 65 DR. I have recently replaced all the caps shown in the output section. I have also replaced all those resistors with 1%. The original components had drifted way out.

When i switch the powet tubes, the current imbalance follows the amp. Both the small bias current difference and the driven current imbalance.

[xtian]

If the imbalance follows the tube, we know the tubes are mismatched, right?

[Stevem]

Sounds to me like one side of the primary of the amps OT has a issue if indeed there is no coupling cap leaking D.C to that tube sockets grid.

Unsolder the 3 wires on its primary side and and use a AC wall wart to perform a OT check that you can find on line.

Make sure that the speaker output jack is open for this test out, just plug a cable into the jack with its other end not hooked up.

[amplifryer]

I will check the OT using ACV. I had ohmed the input side previously with each leg showing 220 ohms to the center tap. I know that doesn't tell the whole story.

It seemed to me that the power tubes and beyond were acting as expected given that their grid signal(s) are being driven with such unequal amplitudes. Maybe I have been chasing my (long) tail phase inverter when it is okay?

[amplifryer]

Duplicate post... Doh!

[amplifryer]

If the imbalance follows the tube, we know the tubes are mismatched, right?

Yes. In this case the imbalance stays with the amp. Even the 4mA difference in bias current.

I only have one spare 6V6 right now but will have a new matched pair tomorrow to experiment with.

[GAStan]

You mentioned you have changed most of the capacitors and resistors in the power section. Did this problem exist prior to changing these out?

[amplifryer]

Yes, the resistors were replaced in an attempt to fix the imbalance. All of the old ones had drifted way out of spec. I started with the 5% tolerance ones then incrementally changed the rest. Some of the DC voltages changed slightly but the output imbalance remained (IMHO) large.

For the purists: This is a "working" amp. I saved all the original components and might put them back in once this problem is resolved. It would be interesting to see how much of an audible difference they make. I also have access to pretty good selection of unused 30 year old carbon comp resistors that have similarly gone way out of spec.

[GAStan]

What does it do when you disconnect the NFB?

[bepone]

here AC output voltage from the PI to the output tubes must be checked, if is the same or one half of PI has less amplification (too much)

[amplifryer]

What does it do when you disconnect the NFB?

That reminds me... when I first starting checking resistors, I discovered the NFB was 82K instead of 820 Ohms. I thought I had found the problem, but replacing it didn't change the imbalance. Disconnected it today and no change in imbalance. I also put in some new "matched" JJ 6V6s. Biased equally @20mA each, but when driven one output rose to 70mA with the other topping out around 30mA.

here AC output voltage from the PI to the output tubes must be checked, if is the same or one half of PI has less amplification (too much)

Yes. Seems to me this is (at least) part of the problem. The AC signal level (on either side of the power tube(s) coupling caps) is not the same. With increasing gain, one PI output grows to twice the amplitude of the other (at least). This drive imbalance appears at the grids of the 6V6s. Wouldn't this cause their outputs to be unequal?

I removed the power tubes and measure a difference in the grid bias voltages. With signal applied, there's a corresponding difference in the AC signal (gain). The pics show the grid DC voltage and grid voltages with signal applied. Scale is 1mS/20V/Div. Bias voltage is set at -40V, about the same as the "most negative" grid.

[bepone]

ok, solution is PI capacitor to the gnd, second grid to the ground is not connected.. somewhere in this line is break (0.1uF-200V)